NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

National Collaborating Centre for Women's and Children's Health (UK). Atopic Eczema in Children: Management of Atopic Eczema in Children from Birth up to the Age of 12 Years. London: RCOG Press; 2007 Dec. (NICE Clinical Guidelines, No. 57.)

Cover of Atopic Eczema in Children

Atopic Eczema in Children: Management of Atopic Eczema in Children from Birth up to the Age of 12 Years.

Show details

6Identification and management of trigger factors

The issues considered in this chapter are potential triggers for atopic eczema in children, clinical methods for identifying trigger factors that exacerbate established atopic eczema in children, and the evidence in relation to avoidance or elimination of potential triggers as part of the management of established atopic eczema in children.

6.1. Potential trigger factors

Studies considered in this chapter

Several reviews have documented factors that are believed to trigger atopic eczema. Trigger factors noted in the reviews are listed here.

Overview of available evidence

Many different factors have been proposed as triggers for atopic eczema in children, mainly as a result of epidemiological studies in which exposure to one or more of the factors has been shown to be associated with increased incidence of atopic eczema and/or exacerbation of established atopic eczema. Potential trigger factors include the following.114,151–159

  • irritants – wool or synthetic clothing, soaps, detergents, perspiration, disinfectants and topical antimicrobials, and many chemical reagents
  • contact allergens – preservatives in topical medications, perfume-based products, metals and latex
  • foods/dietary factors – cow’s milk, eggs, peanuts, tree nuts, wheat, soya, fish, shellfish and (rarely) others such as sesame, kiwi and legumes
  • inhalant allergens (aero-allergens) – house dust mites (Dermatophagoides pteronyssinus and D. farinae), animal dander, cockroach, tree and grass pollens, and moulds
  • microbial colonisation and/or infection – Staphylococcus aureus, streptococcus species, Candida albicans, Pityrosporum yeasts, herpes simplex (colonisation and infection associated with atopic eczema in children is considered separately in Section 7.6)
  • climate – extremes of temperature and humidity, and seasonal variation in the pattern of atopic eczema
  • environmental factors – hard water, cooking with gas, proximity to road traffic, and environmental tobacco smoke
  • familial factors – genetics, family size and sibling order
  • social class (higher incidence in more affluent social classes)
  • concurrent illness and disruption to family life – teething, psychological stress and lack of sleep.

While most triggers lead to reactions confined to the skin, allergic triggers are capable of inducing both skin and systemic responses. These responses are largely mediated via IgE and T cell responses causing immediate (type 1) and/or delayed (late-phase or type 4) allergic reactions. Immediate reactions in the skin can lead to erythema and itching, the onset of urticaria (hives) and/or angioedema (swelling) resulting in an acute flare of atopic eczema. These reactions may be accompanied by systemic features involving the gut (oral pruritus, vomiting, diarrhoea and/or abdominal pain), the respiratory tract (rhinitis, wheeze, cough and stridor (difficulty breathing)) or the cardiovascular system (drop in blood pressure and/or collapse). The involvement of breathing difficulties or a drop in blood pressure constitutes an anaphylactic reaction. Delayed reactions in the skin cause itching and flares of atopic eczema and they may be accompanied by symptoms in the gut (vomiting and/or diarrhoea).

6.2. Identification of trigger factors

Studies considered in this section

Studies evaluating the accuracy of challenge tests (skin tests (skin prick tests and atopy patch tests), IgE tests and skin application food tests (SAFTs)) for the identification of trigger factors for atopic eczema were considered for this section. Skin prick (or puncture) tests are used to detect skin responses to material (for example, foods or inhalant allergens) applied directly to the skin; the responses are usually evaluated over a short period of time (15–20 minutes). The presence of antigen-specific IgE produces a wheal and flare response. The atopy patch test is a skin test where whole food proteins are applied to the skin under occlusion for 24 hours. The test site is evaluated at the time of removal and 48 hours later for evidence of inflammation that can be scored by severity. Controls are applied to determine possible irritant reactions. Raised IgE levels in the blood are an indication of allergy. Other forms of patch tests are used to diagnose contact allergies: the diagnosis and management of contact allergy is outside the scope of this guideline, although such allergies may occur in association with atopic eczema (for example, when a child with atopic eczema develops an irritant reaction or allergy to a topical treatment; see Chapter 7).

Overview of available evidence

No studies have considered the accuracy of any tests for diagnosing inhalant allergies. No tests exist for investigating reactions to climatic, psychological or environmental trigger factors.

Nineteen studies have considered the diagnostic accuracy of one or more tests (skin prick test, atopy patch test, SAFT and/or specific IgE) for detecting food allergy in children with atopic eczema. The double-blind placebo-controlled food challenge (DBPCFC) test is considered to be the gold standard for the diagnosis of food hypersensitivity.160 The reference standard against which the tests were compared was a DBPCFC in eight studies (total n = 787),161–170 and an open food challenge in ten studies (total n = 891).171–180 A further study (n = 437) was designed to use the DBPCFC, but open food challenges were used in children younger than 1 year with a history of immediate reactions.181 A further 11 studies considered how diagnostic accuracy might change when tests were undertaken in different ways, such as using different foods or changing the thresholds for what constituted a positive test. The findings of these studies are described briefly below; more detailed descriptions for each study are presented in Appendix C. Note that some studies were described in more than one publication.

Nine studies present the outcome of food challenge tests.

Identifying food allergy in children with atopic eczema

The studies were heterogeneous in terms of the age of the population evaluated, whether single or multiple tests were evaluated, and in how the tests were undertaken (including variation in the foods tested and which preparation of a particular food was used).

In most of the studies, the age of the population was within the range of 2 months to 12 years. However, older children and adults were included in some studies (up to the age of 28 years).

The foods investigated were predominantly cow’s milk and/or egg, and also wheat, soya, fish and peanuts. Some studies considered diagnostic accuracy for one food only, while other considered accuracy for a range of foods. There was also variation in whether studies reported the diagnostic accuracy for an immediate reaction (usually occurring within 2 hours), a delayed reaction (occurring within 2–72 hours), or any reaction (immediate or delayed, combined). When considering whether a food allergen triggers atopic eczema, delayed reactions are more relevant. Only the minority of studies considered delayed reactions.

The studies were generally consistent in the definition of a positive test (erythema usually with infiltration for an atopy patch test, and a minimum wheal size of 3 mm in diameter on skin prick test). However, while the specific IgE level indicative of a positive test was 0.35 ku/l in all DBPCFC studies, there was greater variability in the level compared with an open challenge (ranging from 0.35 to 99 ku/l).

While all studies that used open challenges were considered to be of poor quality, [EL = DS III] some of the DBPCFC tests were of better quality.162–164,166 [EL = DS II] In most studies it was not clear whether the challenge testing was undertaken blind to (without knowledge of) the results of the tests being evaluated. Neither was it explicitly stated in several studies whether the population evaluated had atopic eczema that was suspected to be exacerbated by food allergy.

As indicated above, because there was heterogeneity in the design of the individual studies, interpretation of the results was difficult. Sensitivity and specificity values were focused on for the main guideline text (although other values such as PPVs and NPVs are reported in Appendix C) as these parameters reflect the performance of the tests, and do not vary with prevalence (unlike predictive values).

Overall summaries of the sensitivity and specificity of the tests for diagnosing reactions to foods (across all studies) are presented below. However, it should be noted that some of these data represent results from only one study.

Atopy patch test

Compared with the DBPCFC test, the atopy patch test (erythema usually with infiltration) had high specificity (81–96%) for any reaction (immediate, delayed, or immediate and delayed combined) to cow’s milk, egg and soya. Specificity for any reaction to wheat was more variable (35–94%). Compared with an open food challenge, the specificity results for any reaction were more variable for cow’s milk, egg, wheat and peanut. Compared with DBPCFC or an open food challenge, sensitivity results for any reaction to a single food (cow’s milk, egg, wheat, soya and peanut) were more variable. Sensitivity and specificity results compared with DBPCFC were both more variable when considered for several foods together (no data compared with an open food challenge).

Skin prick test

Compared with the DBPCFC test the skin prick test (wheal size 3 mm or greater) had high sensitivity (90–95%) for diagnosing an immediate response to fish and peanut, or to several foods together (results from one study); specificity results for these foods were more variable. The sensitivity and specificity for detecting any reaction (immediate, delayed, or combined and compared with DBPCFC) to all other allergens tested (cow’s milk, wheat and soya) were more wide-ranging across studies. Compared with an open food challenge, sensitivity and specificity results for any reaction (immediate, delayed, or combined) to all allergens were also more variable.

Specific IgE

The sensitivity of specific IgE (more than 0.35 ku/l) for detecting any reaction (immediate, delayed, or combined) to cow’s milk and egg was high (83–100%) compared with the DBPCFC test. Sensitivity for detecting an immediate reaction to wheat, soya, fish and peanut compared with DBPCFC was also high (94–97%; one study only). Sensitivity for a combined immediate and delayed reaction to wheat or soya was more variable (no data for delayed reactions). Specificity results for each of the allergens alone or when considered together were more variable. Compared with an open food challenge, both sensitivity and specificity results were less consistent across all foods tested. The specific IgE level indicative of a positive test ranged from 0.35–99 ku/l in the open challenge studies.

Effect of changing test parameters

The available data for each type of test do not show consistency in sensitivity or specificity results. This might reflect the way the particular tests were undertaken or the criteria used to define positive test results. Several studies have considered whether changing certain parameters of a test affects their diagnostic accuracy in children with atopic eczema.165,167,175,182–191 The accuracy of the atopy patch test varied according to the size of the chamber used for occlusion, the vehicle and concentration used to apply the allergen to the skin, and according to which skin sign was taken to indicate a positive test. [EL = 3/EL = DS III] There was some evidence that increasing the wheal size that constituted a positive test on skin prick testing increased the specificity of the test. The specific IgE levels that gave PPVs of 95% for certain allergens were estimated in one study. [EL = DS III]

Outcome of challenge tests

One study recruited 63 children (age range 0.4–19.4 years) with persistent atopic eczema from a dermatology clinic. After screening for food allergy, 19 children underwent DBPCFC. There were four positive challenges to egg, three to milk, three to wheat, two to barley and one to beef. Thirty-seven percent of children in this study were diagnosed with food allergy based on double-blind and open challenges and convincing histories of reactions to foods.143

Three studies conducted in a single clinic reviewed patient records of consecutive referrals for food allergy investigation in children with atopic eczema. In the first study, 107 children (age range 5 months to 12 years, median age 21 months) underwent DBPCFC. Eighty-seven children (81%) showed reactions to at least one food. Early skin reactions were mainly caused by egg and cow’s milk, with wheat and soya producing a small number of reactions. Isolated late reactions were produced by all four allergens in 25% of the total number of reactions.166

The second study reviewed 139 children (age range 2 months to 11.2 years, median age 13 months). DBPCFC was conducted in all children and, of 208 challenges, 111 were positive. Positive challenges were due to cow’s milk (47%), egg (34%) and wheat (19%). It is not clear how many children had at least one positive test. Immediate and late reactions were observed.547

There were 98 children in the third study (age range 2 months to 11.2 years, median age 13 months). Of 173 challenges, 55% were positive, including 45/71 positive challenges to cow’s milk, 28/42 to egg, 18/35 to wheat and 4/25 to soya. It is not clear how many children had at least one positive result. All late reactions (25% of the total) were skin related.165

In one small study, 26 children (age range 16 months to 19 years, median age 11 years) with atopic eczema, elevated IgE level), suspected food allergy and ability to cooperate with food challenge procedures were subjected to DBPCFC. Fifteen children had positive reactions to at least one food. Foods provoking cutaneous reactions were wheat, soya, milk, egg, rye, chocolate and chicken.548

A later study by the same group investigated 113 people (age range 4 months to 24.5 years, median age 6 years) with severe atopic eczema for food allergy. Three hundred and seventy DBPCFCs were undertaken, of which 101 were positive in 56% of patients. Major allergens in this study were egg and peanut, with 85 challenges provoking skin reactions. Allergens provoking a small number (fewer than 6) of reactions were milk, soya, wheat, fish, chicken, pork, beef and potato.549

In another study, food allergy was identified by DBPCFC in 39% (n = 165) of patients (age range 4 months to 21.9 years, mean age 49 months) with atopic eczema attending a specialist allergy clinic. The main foods provoking positive challenges within 2 hours were peanut (n = 27), egg (n = 33) and milk (n = 14). No delayed reactions were observed.550

Sixty-four children (age range 1–10 years, median age 2 years) with atopic eczema and suspected food allergy were investigated in this study. There were 106 DBPCFCs, of which 46% were positive. The allergens tested were cow’s milk, egg, wheat and soya. All of the isolated late reactions (12% of total) were eczematous.169

Seventy-four children (age range 6 months to 16.3 years, median age 2.5 years) with atopic eczema who had been referred to paediatric dermatology or allergy clinics were enrolled for evaluation of food allergy. Six children underwent DBPCFC and positive reactions were to milk (three challenges) and wheat (two challenges).142

6.3. Management of trigger factors

Studies considered in this section

For this section, RCTs evaluating the effectiveness of trigger factor management strategies in children with atopic eczema were considered where available. Where RCTs were not available, studies of any design were considered.

The management of trigger factors in atopic eczema was considered in three systematic reviews.26,156,192 Because two of the reviews included children and adults,26,156 and because of overlap in the studies included in reviews, studies including the population of relevance to this guideline are reported individually here, together with other evidence identified.

Overview of available evidence

The evidence identified in relation to managing trigger factors consisted broadly of exclusion diets and inhalant-allergen avoidance strategies (predominantly avoidance of house dust mite). Various diets have been evaluated, including exclusion of cow’s milk and/or egg, the use of restrictive diets ranging from elemental diets (consisting of products containing amino acids only) to diets including up to 20 foods. Sodium cromoglicate has been evaluated in comparison with, and in addition to, dietary interventions. Probiotics have been evaluated as an adjunct to milk substitutes, and vitamin E and zinc as treatments for atopic eczema.

No evidence was identified regarding avoidance or elimination of the following factors: skin irritants, extremes of temperature or humidity, and stress.

Cow’s milk and egg exclusion diets

Two double-blind randomised crossover trials of egg and cow’s milk exclusion diets involved children with atopic eczema.193,194 [EL = 1-] The studies had 4 or 6 week treatment periods, with a washout period of the same duration in-between. As well as eliminating eggs and cow’s milk, chicken and beef were eliminated, and a soya-based milk substitute given; the control group received a preparation containing a mixture of dried eggs and cow’s milk as a milk substitute. Neither study stated whether there was clear evidence of allergy or intolerance to the eliminated foods, although it was reported in one that three of the 20 children who completed the study had a history of exacerbation of skin symptoms following ingestion of eggs or cow’s milk.193 The most common reason for withdrawal from both studies was non-adherence to the diet. Both studies analysed results only for those who completed treatment.

The first RCT (n = 36; 56% completed), in children aged 2–8 years, found significantly greater improvements in the diet group versus control in atopic eczema activity (global improvement) and skin area affected, sleeplessness and antihistamine usage, with no significant difference between diet and control groups in pruritus (mean improvement 4.49, standard error 2.25).193 The response in the diet group was significantly greater during the first treatment period than the second treatment period for activity, area and sleeplessness, but there was no significant difference between the first and second treatment periods for pruritus or antihistamine usage. For pruritus and sleeplessness this ‘order’ effect was greater than the difference between diet and control groups. It was also reported that there was no correlation between positive prick test to the egg and cow’s milk antigens and response to diet, but no data were reported.193

The second RCT, in children and adults aged 1–23 years, found no significant differences between elimination and control diets in area (mean −1, 95% CI −6 to 3.4) or itch (mean 15, 95% CI −21 to 51) scores. Use of topical corticosteroids was higher during the elimination diet (n = 53; 40 completed).194

Two case series also reported the effects of egg and/or cow’s milk exclusion diets in children with atopic eczema.195,196 One series eliminated cow’s milk and egg from the diet of children (aged 0.4–15 years) for 3 weeks (n = 91; 73% completed and analysed). Improvements in severity scores were reported at endpoint.196 In the other series, children aged 2–14 years who had not responded to usual treatments eliminated cow’s milk and egg, or cow’s milk only, from their diet for 4 weeks. The decision on whether to exclude milk alone or both foods was dependent on which was suspected of precipitating the atopic eczema. However, the outcome was only reported as cure or improvement, with no definition of either term. Additionally it was not clear how many of the children eliminated only milk or both foods from their diet (n = 59).195 [EL = 3]

One case series of children with atopic eczema (n = 11, median age 4 years) documented acute allergic reactions to cow’s milk after prolonged cow’s milk elimination diets.197 [EL = 3]

Egg exclusion diets

Two controlled trials considered the effects of egg exclusion on atopic eczema in infants.198,199 The first was a double-blind RCT in which all the infants had a raised IgE to egg on a radio-allergosorbent test (RAST), and the majority also had a positive test on a DBPCFC test (n = 62; 89% analysed).198 The control group were not given any specific dietary advice. After 4 weeks’ intervention, the reduction in body surface area affected was significantly greater in the diet group compared with control (mean difference 5.25%, 95% CI 0.1% to 10.9%, P = 0.04). Differences between groups in severity scores were not significant (6.1, 95% CI −0.1 to 12.3, P = 0.05), although there were some discrepancies in the trial report between data presented in the text and in the abstract.198 [EL = 1−]

The second trial, described as a single-blind controlled study, reported the proportions of children in four age categories whose condition was ‘better’ after 2 weeks’ treatment. However, ‘better’ was not defined (n = 213; 65% of whom completed treatment and were analysed).This study was not considered further.199 [EL = 1−]

Cow’s milk substitutes

One RCT compared two milk substitutes in infants with atopic eczema and allergy to cow’s milk (shown on DBPCFC; n = 73).200 An amino acid-based formula was compared with a hydrolysed whey formula. Energy intake was similar in both groups. A significant improvement in the SCORAD severity index was seen overall, from a mean of 24.6 at entry to 10.7 after 6 months (P < 0.0001); data were not reported separately by treatment group. In the amino acid group there was a significant increase in the length standard deviation score (SDS) from baseline (P < 0.04), while there was no statistically significant change in the hydrolysed whey group. Weight-for-length values were ‘stable’ in both groups.200 [EL = 1−]

In a randomised study infants with atopic eczema and proven allergy to cow’s milk (on double-blind food challenge) were given hydrolysed whey or amino acid formulas as milk substitutes (n = 45).201 Although the study was described as randomised in the abstract, randomisation was not mentioned elsewhere in the paper. Other dietary restrictions (egg and cereals) were also used in two-thirds of infants. At 8 months, SCORAD scores had improved significantly from baseline in those receiving either milk substitute. The statistical significance of changes in weight and length of infants was also reported, although the data were only presented in graphs. The graphs showed that weight and length increased in both groups in the first month of treatment, and they continued to increase in the amino acid group over the 9 month follow-up period. The pattern in the whey substitute group was less consistent, but weight and length at 9 months appeared to be the same or worse than at baseline. There was overlap of the 95% CIs for the groups for weight indicating that the difference between the groups was not statistically significant for this outcome. However, the difference between groups for length was statistically significant.201 [EL = 1−]

Milk substitutes for women who are breastfeeding

One double-blind crossover RCT considered the effects of an exclusion diet plus a milk substitute in mothers of breastfed infants with atopic eczema (n = 19; 17 completed and analysed; aged 6 weeks to 6 months).202 [EL = 1−] This was the only study relevant to the guideline clinical question in a review of maternal dietary antigen avoidance during pregnancy and/or lactation.203 The foods excluded from the mothers’ diet were cow’s milk, egg, chocolate, wheat, nuts, fish, beef, chicken, citrus fruits, colourings and preservatives. The milk substitutes taken were a preparation containing soya and one containing cow’s milk and egg powder. Area and activity scores (the latter a measure of the severity of the condition on 20 body surface areas) fell from baseline with both milk substitutes after 4 weeks’ use. The difference between groups was not statistically significant (activity score, exclusion diet plus soya versus exclusion diet plus cow’s milk and egg, 104 versus 12.6, P value not reported; area score, exclusion diet plus soya versus exclusion diet plus cow’s milk and egg, 9.0 versus 8.9, P value not reported). A subsequent open, uncontrolled study was undertaken in the same group because of concerns that the soya preparation may have triggered symptoms in the first study (n = 18). In this open study, mothers took their usual diet (containing cow’s milk and egg) for 2 weeks, followed by an exclusion diet for 2 weeks (the same as that taken during the first study), then the usual diet repeated for 2 weeks. Activity and area scores fell significantly after the exclusion diet (at week 4), and remained at around this level after the reintroduction of the usual diet (week 6).202

Restrictive diets

Three studies considered the effects of restricting foods consumed by children with atopic eczema (two case series and one controlled study).204–206 The first case series considered a 2 week diet consisting of up to 19 foods (including meats, carrots, lettuce, parsley, pears, rice, plain flour, sugar, golden syrup, honey, oils, vinegar, salt and pepper and coffee; n = 29, age range 2–12 years).204 The withdrawal rate was 55%, and half of the withdrawals were because the diet was considered to be too restrictive. Thirteen children were evaluated at the end of the 2 week diet. According to the parents’ global assessment seven were improved, three remained the same and three deteriorated. Based on the dermatologist’s assessment of inflammation, lichenification, and cracking, five were improved, seven remained the same and one deteriorated.204 [EL = 3]

The second case series included children aged 10 months to 4 years with severe atopic eczema that persisted despite usual treatment and elimination of the food items to which the child was suspected to be allergic (n = 13).206 A diet consisting of the following foods was taken for 1 month: casein hydrolysate, lamb, rice, corn, corn oil, potato, cucumber, melon, bilberries, salt, sugar, and gluten- and milk-free bread. The numbers of children whose condition improved according to investigator’s and parents’ scores of the severity of the condition were six and eight, respectively. Not all the children who improved according to the investigator improved according to the parents.206 [EL = 3]

The controlled study reported changes in IgE and peripheral blood mononuclear cell concentrations following elimination diets (eliminating the ‘offending foods’) in children aged 3 months to 13 years (n = 153). Changes in severity from baseline were also reported, but a lack of between-group analysis and of details of the diets given made interpretation difficult.205 [EL = 2−]

Few foods diets

Three studies considered the effectiveness of ‘few foods’ diets (eliminating all but five to eight foods); these were a single-blind RCT and two case series.207–211 The single-blind RCT evaluated a diet (including either whey or casein hydrolysate milk formula) in children aged 0.3–13 years with atopic eczema that persisted despite conventional treatment (n = 85).207 After 6 weeks there were no significant differences between the diet group and control group (continued usual diet) in changes in any outcome (body surface area affected, severity, daytime itch, or sleep disturbance). The withdrawal rates were 59% in the diet group (the most common reason for withdrawal being non-adherence), and 15% in the control group; only results for those who completed the 6 week intervention period were analysed.207 [EL = 1−]

A case series of children with extensive atopic eczema (affecting 30% or more of body surface area) that responded poorly to conventional treatment or who had a history of food intolerance were given a few (six) foods diet (n = 63, age range 0.4–14.8 years).208,210 After 6 weeks, the median severity score fell by 33%, with 52% having a 20% or greater reduction in score. ‘Little or no benefit’ was seen in 39%. The withdrawal rate was 14%. Of the 68% who were followed up for 1 year, the outcome was similar in children regardless of their response to the 6 week few foods diet, although no data were presented.208 Some children from this study were subsequently given an elemental diet (see below).209,210 [EL = 3]

Another case series of children with severe atopic eczema evaluated a few foods diet (n = 66, age range 0.6–17 years).211 Twenty-four patients (36%) were reported to have ‘worthwhile’ improvement (the term ‘worthwhile’ was not defined) from the diet (median duration 26 days, range 19–44 days). In 15 of these (23% of the total group), improvement persisted on dietary treatment, but three withdrew because the diet was too burdensome. Overall, 12 (18%) persisted with the diet and had continued benefit over the duration of follow-up (mean 48 weeks, range 26–71 weeks). The outcomes beyond this follow-up period were not reported.211 [EL = 3]

Elemental diets

A randomised crossover study in infants and children with a positive skin prick test and raised cow’s milk-specific and soya bean-specific IgE evaluated an amino acid-based elemental diet (n = 15; 11 analysed).212 Dairy- or soya-based products were also excluded from the diet. The control group continued with a pre-existing formula (no further details were reported). Following 6 weeks’ treatment, there were no significant differences between the amino acid-based elemental diet and the control diet in SCORAD scores or in global health scores. A significant treatment-by-period interaction was reported for SCORAD, which was greater than the between-group treatment difference.212 [EL = 1−]

A case series of children aged 0.4–13 years with severe and extensive atopic eczema were hospitalised for treatment with an elemental formula only (the product contained 100% free amino acids). Pet and house dust mite avoidance measures in the children’s homes were a prerequisite for treatment (n = 37).209,210 After a median duration of 30 days’ treatment, 27% were considered to be ‘treatment failures’ because their severity scores were unchanged or worse compared with baseline. In the 73% for whom treatment was considered to be successful, the severity scores decreased to 27% of the baseline score (range 3–67%; no further details of who had greatest or least benefit were presented). No significant differences in demographics or in clinical features were found between those in whom treatment was successful and those in whom it was not successful. Reported adverse effects were weight loss of up to 17% (in 30 of 34 evaluated), loose stools (19%), and a reduction in serum albumin in 93% of 27 children in whom this parameter was measured (from a mean of 30.8 g/l to a mean nadir of 21.2 g/l). No electrolyte disturbances were reported.209,210 [EL = 3]

A further case series reported the outcomes of an elemental diet in children with atopic eczema (n = 10, age not specified).213 Only the elemental diet was used for 2 weeks, followed by addition of pumpkin, potatoes, zucchini, apples, pears and pure vegetable margarine. Two children stopped using the diet after 1 week. In the other eight, the atopic eczema scores (a measure of severity, extent and of treatment required) fell significantly at 6 weeks, and increased again after reintroduction of their usual diet. Adverse effects were not considered. [EL = 3]

The effectiveness of a ‘home-made meat-based formula’ diet was considered in a case series of children with severe atopic eczema (n = 16, aged 5–24 months).214 The children had positive skin prick test results to cow’s milk, egg, and wheat and/or soya. The formula consisted of lamb, olive oil, rice flour and water, supplemented with calcium and vitamin D. After 1 month, the severity score had fallen (no statistical analysis reported), with no significant changes in lipid levels. It was reported that all the children had gained weight normally, but no data were presented.214 [EL = 3]

Sugar exclusion

One study considered whether avoiding sugar had an impact on atopic eczema in children and adults (n = 30; 9 children).215 No significant changes in SCORAD severity scores were seen in the children’s atopic eczema 1 week after the elimination diet, and differences in SCORAD following a double-blind or placebo food challenge were also not significant. Aspartame was offered as a replacement for sugar, but it was not clear how many took this.215 [EL = 3]

Sodium cromoglicate

Four studies evaluated the effectiveness of sodium cromoglicate therapy in children with atopic eczema, either compared with or in addition to an elimination diet (three RCTs and one case series).216–219 The first RCT compared a restricted diet (consisting of 12 foods) with oral sodium cromoglicate in children aged 5 months to 14 years. After 4 weeks’ treatment, there were no significant differences between groups in severity or disease extent (n = 1085, 80% analysed).216 [EL = 1−]

Two placebo-controlled crossover RCTs evaluated the addition of sodium cromoglicate to an elimination diet tailored to individual children with atopic eczema.217,219 In the first RCT, significant improvements in severity were reported for both groups after 6 weeks’ treatment, but no between-group analysis was reported to allow comparison between groups (n = 29, 76% completed and analysed, aged 3–12 years).217 [EL = 1−]

The second crossover RCT found no significant differences between investigators’ or parents’ assessments of severity at 8 weeks when treatment with sodium cromoglicate was followed by placebo. However, improvements in severity were significantly greater with sodium cromoglicate when the treatment sequence was reversed (that is, placebo taken first, n = 31, 94% analysed, aged 6 months to 10 years).219 [EL = 1−]

In the case series, sodium cromoglicate was added to an individually tailored exclusion diet in children aged 1–15 years (n = 35). However, the outcome of sodium cromoglicate treatment was only expressed as ‘improved’ or clear/almost clear, with no definitions given. Without a control group the study was of limited value, and it was not considered further.218

Vitamin and mineral supplementation

Two placebo-controlled RCTs considered the effectiveness of zinc or vitamin E for atopic eczema.220,221 The trial involving zinc included children aged 1–16 years who continued with their usual treatments for atopic eczema (emollients and topical corticosteroids). Itch scores were significantly higher in children treated with zinc than with placebo, otherwise there were no significant differences in any outcome at 8 weeks (sleep disturbance, redness, surface area or combined disease severity scores, or in use of other treatments; n = 50, 84% analysed).220 [EL = 1−]

The trial of vitamin E included children and adults (n = 96, aged 10–60 years).221 Treatment with emollients was continued. Vitamin E or placebo was given for 8 months, after which the global assessment of the condition (classifications not defined) found worsening in 8% of the vitamin E group versus 78% in the placebo group; no change in 12% versus 11%, slight improvement in 20% versus 9%, great improvement in 46% versus 2%, and almost complete remission in 14% versus 0%. No statistical analysis of the data was presented and no adverse effects were reported.221 [EL = 1−]


Three double-blind RCTs considered the effectiveness of a milk substitute supplemented with probiotics for the treatment of atopic eczema in infants with suspected cow’s milk allergy.222–224 The cow’s milk substitute in all three studies was a hydrolysed whey formula, with Lactobacillus added in the intervention group. Two studies had an additional intervention group: one received a mixture of probiotics (Lactobacillus, Bifidobacterium, Propionibacterium) and the other received Lactobacillus rhamnosus. Control groups received the hydrolysed whey formula only. Two studies evaluated 1 month’s use. Of these, the study with three treatment arms found no significant differences between the groups treated with probiotics and the control group in changes in SCORAD severity scores (n = 252, 91% completed and analysed).222 [EL = 1−] The second study reported significant improvements in SCORAD scores from baseline in the group receiving the hydrolysate plus probiotic. However, no between-group analysis was reported (n = 31).223 [EL = 1−] The treatment period in the remaining study was 3 months and no differences in SCORAD reduction were found between the three groups.224 [EL = 1−]

House dust mite avoidance

Two RCTs in children225,226 and one involving children and adults227 considered the effectiveness of house dust mite avoidance. One of the RCTs evaluated the effects of bedding encasement with microfine fibres on mite sensitisation in children with atopic eczema, but did not report any clinical outcomes (only IgE and house dust mite levels were measured) and is therefore not considered further (n = 57).225

A 2 month placebo-controlled RCT in young children (aged 2–10 years, mean 3.9 years) evaluated house dust mite allergen avoidance measures. The children had moderate atopic eczema (SCORAD 27–33) associated with high total and/or specific IgE serum levels (n = 41).226 The mite avoidance measures consisted of encasing mattresses and pillows, a hot weekly wash of bedding, vacuuming of living rooms and bedrooms at least twice a week, and removing or washing soft toys once a week; pets were not allowed. In the control group the previous house cleaning strategy was continued. After 2 months’ intervention, a significant reduction in the SCORAD index was reported in the avoidance group; the score also fell in the control group, but no between-group analysis was reported. Significant reductions from baseline in dust load and house dust mite allergen concentrations were reported in the avoidance group, but not in the control group; again no between-group analysis was reported, nor was there any consideration of whether groups were similar at baseline in the parameters measured.226 [EL = 1−]

A further double-blind RCT (n = 60, aged 7–65 years) compared a house dust mite avoidance strategy (GORE-TEX® bedding system, carpet spraying and use of a high-filtration vacuum cleaner) with placebo in children and adults who had positive results in skin prick tests using a range of inhalant allergens. 227 After 6 months, the reduction in severity (measured using SASSAD) was significantly greater in the avoidance group compared with placebo (mean difference 4.2, 95% CI 1.7 to 6.7, P = 0.008; mean difference in final severity score in children (aged younger than 17 years) 11.1, 95% CI −3.1 to 25.3, P = 0.019). The reduction in bed mattress dust load was significantly greater in the intervention group compared with placebo (98% versus 16%, P = 0.002). Median reductions in the concentrations of the house dust mite allergen in bedroom or living room carpets were not significantly different between intervention and control groups (91% versus 89%, P = 0.94 and 76% versus 38%, P = 0.27, respectively).227 [EL = 1−]

In a non-randomised controlled study, the effectiveness of an air cleaning system (in a ‘clean-room’) for the treatment of people aged 8–75 years with atopic eczema who had high specific IgE levels to house dust mite was evaluated (n = 30).228 Participants were hospitalised for 3–4 weeks, and were exposed to either an air cleaning system in a clean-room, or to a similar room without the air cleaning system. The only clinical outcome reported was time to recurrence of symptoms – it was unclear whether this referred to all symptoms or specifically to itchiness. It was reported that time to recurrence of symptoms in those in the clean room who had high IgE to house dust mite was a mean of 8.4 months, whereas in those with no raised IgE to house dust mite the time to recurrence was 1.7 months. In the control group (no air filtration system, and high IgE to house dust mite) the time to recurrence was 1.6 months. No baseline data were reported.228 [EL = 2−]

Hyposensitisation to house dust mite

Two studies considered the effects of hyposensitisation to house dust mite on atopic eczema in children who had a positive skin prick test result to this allergen. One was a double-blind RCT with 6 months’ follow-up (n = 26),229 [EL = 1−], and the second was a controlled trial of up to 3 years’ duration (n = 60).230 [EL = 2−] Neither study found significant differences in the severity or clinical features of atopic eczema between those receiving hyposensitisation therapy and those in the control groups (placebo or continued usual treatment).229,230

Evidence statement for identification and management of trigger factors

Potential trigger factors

A plethora of potential triggering factors for atopic eczema has been documented in the scientific literature, including irritants, contact allergens, food and dietary factors, inhalant allergens, microbial colonisation of skin, climate, environmental factors and familial factors. Many of these have been considered only in the context of primary causes/prevention of atopic eczema (which are outside the scope of this guideline), rather than in terms of triggering exacerbations of established atopic eczema. Most data in relation to the identification and management of trigger factors relate to testing for food allergies and elimination diets, and avoidance strategies for inhalant allergens.

Identification of trigger factors

There has been little consistency among the studies that have considered the accuracy of atopy patch tests, skin prick tests and specific IgE for identifying food allergy in children with atopic eczema. The studies varied in the age of the study populations, the foods tested, the standard against which results were compared (DBPCFC or open food challenge), and in the way the tests were undertaken (the types of foods used and the criteria used to define positive test results). There was evidence that changing the definition of a positive test result for the atopy patch test, the skin prick test and specific IgE changed the diagnostic accuracy of the tests. [EL = DS III]

Only a minority of studies focused on delayed reactions (in which the suspected food caused exacerbation of atopic eczema). The studies varied in whether they reported diagnostic accuracy of a test for a specific allergen or for all allergens together, and whether they considered accuracy for detecting immediate and/or delayed reactions.

The general trends for sensitivity and specificity of tests for diagnosing any reaction to foods across these studies were as follows:

  • The atopy patch test (erythema usually with infiltration) had high (more than 80%) specificity for cow’s milk, egg, soya and peanuts compared with DBPCFC or open food challenges. Specificity results for wheat and sensitivity results for all foods were more variable.
  • The skin prick test (wheal size 3 mm or greater) had high sensitivity for egg, fish and peanut compared with DBPCFC; results for cow’s milk, wheat and soya were variable. Sensitivity results compared with open challenge were more variable. Specificity results for all allergens were more variable.
  • The sensitivity of specific IgE for cow’s milk and egg was high compared with DBPCFC, but less consistent compared with open food challenge. Specificity results for wheat were more variable. Sensitivity results for all foods were more variable. The specific IgE level indicative of a positive test result was 0.35 ku/l in DBPCFC studies, but ranged from 0.35 to 99 ku/l in the open challenge studies.

Studies that reported the diagnostic accuracy of a test for any food allergen might have been useful for ruling out food allergy, but the available data did not show consistency in sensitivity or specificity results. [EL = DS III]

Outcome of challenge tests

Positive challenges were reported in 40–81% of oral food challenge tests. Egg, cow’s milk and nuts were consistently identified as being the most frequent allergens to trigger an immediate response. Wheat, soya, fish and shellfish were also identified as additional food allergens triggering immediate or delayed responses.

Immediate reactions involved the skin, gut and respiratory systems. Skin reactions were reported as eczematous symptoms or urticaria. Late eczematous reactions occurred in 45% of challenges in two studies. One study reported only a delayed reaction (that is, no immediate response) in 12% of children.

The prevalence of food allergy in children with atopic eczema in secondary care settings was estimated to be 37–56%. [EL = 2]

Management of trigger factors

Most evidence regarding the management of trigger factors in children with atopic eczema related to dietary exclusions or house dust mite avoidance strategies. There was little consistency across studies in the type of diet evaluated, and indications for special diets were not always made clear. There were confounding factors in many studies, for example exclusion of other foods in addition to cow’s milk and egg in studies specifically evaluating exclusion of cow’s milk and egg.

In crossover RCTs, 4–6 weeks’ cow’s milk exclusion diets produced conflicting results with significant differences between treatment and control arms in some, but not all, outcomes. The most common reason for withdrawal from the studies was non-adherence to the diet. [EL = 1−] In infants with moderate to severe eczema and cow’s milk allergy, those fed a whey formula did not exhibit a linear growth pattern during the 9 month follow-up period whereas those fed an amino acid formula did. [EL = 2−]

Egg exclusion alone in children with suspected egg allergy led to improvements in extent, but not severity, of atopic eczema (one RCT). [EL = 1−]

There was no good evidence to support the use of the following interventions in the management of children with atopic eczema: ‘few foods’ diets, elemental diets, addition of probiotics to milk substitutes, sodium cromoglicate (alone or in addition to restricted diets), or excluding foods from the diet of women who are breastfeeding. [EL = 1−]

There was some evidence that house dust mite avoidance strategies in children and adults led to greater improvements in atopic eczema severity than placebo after 2–6 months. [EL = 1−]


There was no published evidence on the cost-effectiveness of any of the tests for diagnosing trigger factors. A cost-effectiveness model to assess the comparative advantage of alternative means of diagnosing trigger factors was not feasible owing to the complexity of the data required (which would require assessment of all the consequences of true and false positive and negative diagnoses of a range of trigger factors on the management and subsequent outcomes of atopic eczema in children) and was not identified as a priority for this guideline.

From evidence to recommendations

It is the GDG’s view that a clinical assessment (clinical history and physical examination) should play a key role in identifying potential trigger factors, including suspected food allergy. The clinical pattern of atopic eczema can indicate potential allergies (particularly to inhalant allergens).

Food allergy plays an important role in triggering both immediate and delayed skin reactions in children with moderate or severe atopic eczema. The prevalence in the community is currently unknown but figures for children reviewed in secondary care settings range from 37% to 56%. The main foods triggering immediate reactions are cow’s milk, egg and nuts. Immediate reactions to wheat, soya, fish and shellfish occur less frequently.

The child’s age should be considered during history taking. Parents should be questioned about the pattern of atopic eczema in the child from birth. Allergy to cow’s milk, egg and soya is less likely if atopic eczema developed after 2 years of age. History taking should include consideration of foods eaten, quantities (how much and how often), and foods not eaten in order to direct which foods to test for. The GDG believes that the following are signs of an immediate allergic reaction to food, although evidence was not specifically sought to assess this: widespread redness or rash, urticaria, increased itching, facial swelling, rhinitis, wheeze, cough, difficulty breathing, vomiting, abdominal pain, voice change, profound drowsiness, floppiness and/or loss of consciousness.

It is the GDG’s view that children with atopic eczema who are suspected of having a food allergy should be referred for specialist investigation and management of the allergy. Owing to the heterogeneity of published diagnostic accuracy studies and the relative lack of data on costs for, or effectiveness of, tests for specific allergens in the age groups in which food testing is most likely to be required, the GDG felt unable to recommend any test for ruling out allergy. The 95% PPVs for some tests for different food allergens have been estimated in populations outside the UK and it is not certain whether these data are transferable to the UK population. Therefore, none of the tests can be used to rule in allergy and so the DBPCFC test remains the gold standard test for diagnosing food allergy.

For bottle-fed babies who are suspected of having a food allergy, the GDG consensus was that a trial of extensively hydrolysed formula milk should be offered. Although some European countries restrict this to casein-based formulas because they are thought to be less allergenic, there are only two such formulas on the market in the UK and the GDG did not consider there to be enough evidence of clinical or cost-effectiveness to justify this restriction in the NHS. Amino acid formulas are possibly better than casein-based formulas for promoting normal growth, but they are more expensive and they have not been demonstrated to be more cost-effective.

Goat’s milk should not be offered to bottle-fed babies because it is nutritionally inadequate and shares 95% of cross-reacting allergens with cow’s milk. Soya-based formulas contain phytooestrogens and are not recommended in the UK as the primary protein source in infants under 10 months. The GDG also considered that peanut allergy was more likely to develop if soya milk was consumed.

The GDG found little evidence to assess the effectiveness of elimination diets for breastfeeding mothers of children with atopic eczema. There was some support within the group for recommending elimination diets, but these were not already common practice in the NHS. The majority decision of the GDG was that women should be informed that the evidence base for elimination diets is thin but that they can be undertaken under the supervision of a specialist if food allergy is strongly suspected.

The GDG believes that there is not enough evidence to recommend house dust mite elimination measures or removal of pets, although it has been suggested that the timing of exposure to pets may affect the development of allergies. GDG discussion highlighted the possible negative psychological impact of removing pets from children. The house dust mite elimination strategies evaluated in published clinical trials did not reflect current elimination practices. Elimination strategies may not be practicable in many cases.

Although the following potential trigger factors were explicitly mentioned in the guideline scope, the GDG did not find sufficient evidence to evaluate the effectiveness of their avoidance in the management of established atopic eczema: hard water, extremes of temperature or humidity, and stress. Nevertheless, the consensus view of the GDG based on their collective clinical experience was that humidity, stress or extremes of temperature could exacerbate atopic eczema in children and that they should be avoided where possible. The avoidance of irritants contained in topical preparations used to treat atopic eczema is considered in Chapter 7.

The GDG found no evidence which could be used to evaluate allergy testing in children with atopic eczema offered on the high street or over the internet (this could include conventional tests discussed above, and/or analysis of hair samples, Vega testing, etc). The GDG believes that any form of allergy testing outside a recognised clinical setting (such as the NHS) should be discouraged to avoid misinterpretation of results.

Recommendations for identification and management of trigger factors

When clinically assessing children with atopic eczema, healthcare professionals should seek to identify potential trigger factors including:

  • irritants, for example soaps and detergents (including shampoos, bubble baths, shower gels and washing-up liquids)
  • skin infections
  • contact allergens
  • food allergens
  • inhalant allergens.

Healthcare professionals should consider a diagnosis of food allergy in children with atopic eczema who have reacted previously to a food with immediate symptoms, or in infants and young children with moderate or severe atopic eczema that has not been controlled by optimum management, particularly if associated with gut dysmotility (colic, vomiting, altered bowel habit) or failure to thrive.

Healthcare professionals should consider a diagnosis of inhalant allergy in children with seasonal flares of atopic eczema, children with atopic eczema associated with asthma or allergic rhinitis, and children aged 3 years or over with atopic eczema on the face, particularly around the eyes.

Healthcare professionals should consider a diagnosis of allergic contact dermatitis in children with an exacerbation of previously controlled atopic eczema or with reactions to topical treatments.

Healthcare professionals should reassure children with mild atopic eczema and their parents or carers that most children with mild atopic eczema do not need to have tests for allergies.

Healthcare professionals should advise children with atopic eczema and their parents or carers not to undergo high street or internet allergy tests because there is no evidence of their value in the management of atopic eczema.

Healthcare professionals should offer a 6–8 week trial of an extensively hydrolysed protein formula or amino acid formula in place of cow’s milk formula for bottle-fed infants aged under 6 months with moderate or severe atopic eczema that has not been controlled by optimal treatment with emollients and mild topical corticosteroids.

Healthcare professionals should refer children with atopic eczema who follow a cow’s milk-free diet for longer than 8 weeks for specialist dietary advice.

Diets based on unmodified proteins of other species’ milk (for example, goat’s milk, sheep’s milk) or partially hydrolysed formulas should not be used in children with atopic eczema for the management of suspected cow’s milk allergy. Diets including soya protein can be offered to children aged 6 months or over with specialist dietary advice.

Healthcare professionals should inform women who are breastfeeding children with atopic eczema that it is not known whether altering the mother’s diet is effective in reducing the severity of the condition. A trial of an allergen-specific exclusion diet should be considered under dietary supervision if food allergy is strongly suspected.

Healthcare professionals should inform children with atopic eczema and their parents or carers that it is unclear what role factors such as stress, humidity or extremes of temperature have in causing flares of atopic eczema. These factors should be avoided where possible.

Research recommendations for identification and management of trigger factors

How effective and cost-effective is the use of house dust mite avoidance strategies in the treatment of childhood atopic eczema and which strategies, if any, are the most effective?

Why this is important

There are conflicting data on the effectiveness of using house dust mite avoidance strategies in the management of childhood atopic eczema. Many of the currently suggested techniques are time-consuming and expensive for parents/carers and it is important to establish their value.

When and how should children with atopic eczema be tested for allergies (skin prick tests, allergen-specific immunoglobulin E), and how can the diagnostic accuracy and effect on clinical outcomes of the tests be improved?

Why this is important

Parents and carers of children with atopic eczema often ask for allergy testing. However, there is confusion among clinicians about which tests are the most appropriate for different age groups. Interpretation of test results requires training and can be difficult because the diagnostic accuracy is uncertain; carrying out the tests is expensive and time-consuming and requires special training. The research should encompass clinical outcomes (for example, control of atopic eczema) in children who are diagnosed with allergies and undergo interventions to avoid exposure to relevant allergens. The results of the research will enable effective and cost-effective use of NHS resources.

How should exposure to pets be managed in children with atopic eczema; at what age does allergy occur and does tolerance develop?

Why this is importantt

Many children with atopic eczema show signs and symptoms of allergic reactions when in contact with animals such as cats, dogs and horses. However, clinical experience has found that many people report tolerance of their own pet but not others and this tolerance may be lost when teenagers move away from home. In cases of extreme allergy, some practitioners recommend the removal of the pet, while others suggest limited ‘managed’ exposure. There is a single abstract report of children choosing their pet as one of their three most favourite items and the psychological distress of pet removal may not be justified. Clear guidance is needed on the correct management of pet allergy in children with atopic eczema.

What is the optimal feeding regimen in the first year of life for children with established atopic eczema?

Why this is important

Dietary manipulation has the potential to decrease disease severity in children with proven food allergy. A study is needed to explore the potential benefits and harms of delaying the introduction of allergenic foods such as milk, egg and peanuts in infants with early signs of atopic eczema to assess the potential impact on atopic eczema severity and the subsequent development of food allergy, asthma and allergic rhinitis.

Copyright © 2007, National Collaborating Centre for Women’s and Children’s Health.

No part of this publication may be reproduced, stored or transmitted in any form or by any means, without the prior written permission of the publisher or, in the case of reprographic reproduction, in accordance with the terms of licences issued by the Copyright Licensing Agency in the UK []. Enquiries concerning reproduction outside the terms stated here should be sent to the publisher at the UK address printed on this page.

The use of registered names, trademarks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant laws and regulations and therefore for general use.

Bookshelf ID: NBK49357


Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...